IMPACT OF TELEMEDICINE ENCOUNTERS ON SURVIVAL OUTCOMES: A TIME-VARYING COX ANALYSIS USING EHR-DERIVED DATA
Author(s)
Deepika Paratane, MS1, Blythe Adamson, MPH, PhD2, Antal T. Zemplenyi, MSc, PhD3;
1University of Colorado Skaggs School of Pharmacy and Pharmaceutical Sciences, PhD Student, Denver, CO, USA, 2Flatiron Health, New York, NY, USA, 3University of Colorado Skaggs School of Pharmacy and Pharmaceutical Science, Denver, CO, USA
1University of Colorado Skaggs School of Pharmacy and Pharmaceutical Sciences, PhD Student, Denver, CO, USA, 2Flatiron Health, New York, NY, USA, 3University of Colorado Skaggs School of Pharmacy and Pharmaceutical Science, Denver, CO, USA
OBJECTIVES: While RCTs have shown that web-based services improve survival in lung cancer patients, real-world evidence on impact of telemedicine in specific biomarkers like EGFR or ALK remains unclear. Patients with EGFR or ALK-positive NSCLC need ongoing toxicity monitoring on long-term oral targeted therapies. This research aims to assess the association between telemedicine use and progression or death in advanced NSCLC.
METHODS: We conducted a retrospective cohort study of patients with advanced NSCLC initiating 1l therapy(1LT) (i.e., index date) using EHR-derived data from the US Flatiron Health Research Database from 2011 to 2023. Intervals of person-time were created to represent periods before and after each patient's first telemedicine encounter. Extended-Cox Proportional Hazards model was used to model telemedicine as a time-varying exposure while adjusting for demographic, clinical, and time-related covariates. Sensitivity analyses recalculated event dates using start or end of death-month to assess robustness. Additionally, separate adjusted models were used to estimate effect of telemedicine within each therapy group.
RESULTS: In the extended-cox model, telemedicine was associated with a significantly lower hazard of progression or death (n=5,812; HR =0.87, 95% CI: 0.78-0.97,p=0.0013). In exploratory stratified analyses by treatment class, association was directionally consistent for EGFR-TKIs (n=3,809; HR=0.86, 95% CI: 0.75-0.97,p=0.018), ALK-TKIs (n=647; HR=0.84, 95% CI: 0.59-1.19), and chemotherapy (n=933, HR=0.79, 95% CI: 0.59-1.06), with limited precision in smaller subgroups. Known prognostic factors such as ECOG-performance, squamous histology, smoking history, and treatment groups were strong predictors of hazard. Sensitivity analyses using alternate death date-of-month definitions confirmed the findings with start (HR=0.88, 95% CI: 0.79-0.98,p=0.0164) and end (HR=0.86, 95% CI: 0.77-0.96,p=0.0065) dates.
CONCLUSIONS: Telemedicine use during 1LT was consistently associated with reduced risk of progression or death in patients with advanced NSCLC. These findings support the potential value of telemedicine as an addition to routine cancer care.
METHODS: We conducted a retrospective cohort study of patients with advanced NSCLC initiating 1l therapy(1LT) (i.e., index date) using EHR-derived data from the US Flatiron Health Research Database from 2011 to 2023. Intervals of person-time were created to represent periods before and after each patient's first telemedicine encounter. Extended-Cox Proportional Hazards model was used to model telemedicine as a time-varying exposure while adjusting for demographic, clinical, and time-related covariates. Sensitivity analyses recalculated event dates using start or end of death-month to assess robustness. Additionally, separate adjusted models were used to estimate effect of telemedicine within each therapy group.
RESULTS: In the extended-cox model, telemedicine was associated with a significantly lower hazard of progression or death (n=5,812; HR =0.87, 95% CI: 0.78-0.97,p=0.0013). In exploratory stratified analyses by treatment class, association was directionally consistent for EGFR-TKIs (n=3,809; HR=0.86, 95% CI: 0.75-0.97,p=0.018), ALK-TKIs (n=647; HR=0.84, 95% CI: 0.59-1.19), and chemotherapy (n=933, HR=0.79, 95% CI: 0.59-1.06), with limited precision in smaller subgroups. Known prognostic factors such as ECOG-performance, squamous histology, smoking history, and treatment groups were strong predictors of hazard. Sensitivity analyses using alternate death date-of-month definitions confirmed the findings with start (HR=0.88, 95% CI: 0.79-0.98,p=0.0164) and end (HR=0.86, 95% CI: 0.77-0.96,p=0.0065) dates.
CONCLUSIONS: Telemedicine use during 1LT was consistently associated with reduced risk of progression or death in patients with advanced NSCLC. These findings support the potential value of telemedicine as an addition to routine cancer care.
Conference/Value in Health Info
2026-05, ISPOR 2026, Philadelphia, PA, USA
Value in Health, Volume 29, Issue S6
Code
HSD60
Topic
Health Service Delivery & Process of Care
Disease
No Additional Disease & Conditions/Specialized Treatment Areas, SDC: Oncology